Online Issues
<< All Back-issues
<< This Issue's Table of Contents
ILAR Journal V40(4) 1999
Animal Models of Inflammation
Borna Mehrad and Theodore J. Standiford
| Borna Mehrad, M.D., is a Pulmonary Fellow, and Theodore J. Standiford, M.D., is an Associate Professor, in the Division of Pulmonary and Critical Care Medicine at the University of Michigan Medical Center, Ann Arbor, Michigan. |
Introduction
Pneumonia is one of the most common and serious infections afflicting humans. In the United States, pneumonia affects 3 million adults each year, resulting in 500,000 hospital admissions and 45,000 deaths (Anonymous 1997). Pneumonia remains the sixth most common cause of death overall and the most common cause of death from an infectious disease in the United States, and its mortality rate has increased since 1980 (Anonymous 1988; Garibaldi 1985; Pinner and others 1996). The increasing proportion of older individuals in industrialized countries, as well as growing numbers of immunocompromised patients worldwide, has resulted in a large population that is particularly susceptible to pneumonia. In addition, widespread use of antibiotics has led to increasing antibiotic resistance among the causative pathogens, producing infections that are difficult to treat and have a poorer outcome (Doern1995; Holmberg and others 1987).
In this context, modulation of the host immune response as an adjunctive treatment of pneumonia presents a tantalizing prospect. The rationale for this approach stems from clinical observations of increased incidence and severity of pneumonia in diverse populations of immunocompromised hosts (Sanders and Crystal 1997). These observations are confirmed by animal studies that have shown the feasibility of altering the outcome of pneumonia by specific augmentation or depletion of inflammatory mediators. Translation of such interventions in animal models to treatment of patients requires a thorough understanding of the normal host response to various pathogens in pneumonia.
Overview of Host-Pathogen Interactions
Microorganisms that cause pneumonia in humans reach the lung by one of four mechanisms. The vast majority of organisms responsible for community-acquired and nosocomial pneumonia first colonize the pharynx (Johanson and others 1969, 1972). These pathogens then gain entry to the lungs by microaspiration of the oropharyngeal secretions, an event that has been shown to occur frequently in normal adults during sleep (Huxley and others 1978). In contrast, some organisms, including Legionella and Mycobacterium species, enter the lungs directly as inhaled aerosols. Rarely, organisms may reach the lungs by hematogenous spread or direct penetration from the abdomen or chest wall.
Although the mouth and pharynx can be colonized by microorganisms in healthy individuals, the normal pulmonary defenses maintain a sterile environment below the level of the larynx. Pulmonary defense against microbial pathogens is in large part accomplished by physical barriers provided by the upper airways, the cough and epiglottic reflexes, mucociliary clearance, and the antimicrobial properties of airway mucosal surfaces (Reynolds 1997). The organisms that elude these barriers and reach the alveolar units are rapidly opsonized and ingested by resident alveolar macro-phages (Jackson and others 1967). The course of infection beyond this point is determined by properties of both the host and the pathogen, which include the inoculum size, the surface characteristics and virulence of the microorganism, and the ability of the alveolar macrophage to kill the ingested pathogen (Toews 1994). Innate immunity is the principle host defense mechanism in most acute pneumonias. However, defense against some organisms, such as mycobacteria and endemic fungi, requires the involvement of specific immunity, including cell-mediated and antibody-mediated mechanisms.
The principal cells involved in pulmonary innate immune responses are activated alveolar macrophages, recruited neu-trophils, and recruited monocyte/macrophages. These cells orchestrate the immune response by elaborating a complex network of pro- and antiinflammatory messenger molecules, the best studied of which are the cytokines. Cytokines are involved in the immediate response after the recognition of a pathogen, for the influx of immune cells to the site of infection, the activation of resident or recruited cells to ingest and kill the pathogen, and lastly, for the repair of the lung during the resolution of acute injury. It is important to note that many cytokines exhibit pledtropic effects and may be involved in multiple facets of the innate immune response.
Development of Animal Models of Pneumonia
Selection of Experimental Animals
Animal models of pneumonia have been developed in a number of different species. Studies of larger animals, such as sheep and dogs, allow for replication of clinical circumstances and diagnostic tests that apply to patients; examples of the latter include models of ventilator-associated pneumonia (Hanly and Light 1989) and evaluation of radiograms, bronchoscopy, and invasive hemodynamic monitoring in pneumonia (Caidwell and others 1975; Moser and others 1982; Patterson and Todd 1982). Another consideration in the selection of the species is the anatomical and cellular similarity to humans. For example, mouse conducting airways are lined by cuboidal (rather than columnar pseudo-stratified) epithelium, and sheep and pigs have pulmonary intravascular macrophages. In this context, primate models of pneumonia have the additional advantage of greater similarity to human disease (Barile and others 1993; Baskerville and others 1991). All large animal models share the disadvantage of the great expense of obtaining and maintaining the animals. In addition, limited availability of investigative tools, such as neutralizing antibodies, receptor agonists, and transgenic or knockout animals, hampers mechanistic research using these models. Our laboratory uses murine models of pneumonia, taking advantage of their relatively low cost and the availability of reagents and genetically altered animals. This review focuses on the use of murine models of pneumonia, with a special emphasis on techniques to identify the role of specific cytokines as mediators of events that occur in pneumonia.
Preparation of Infectious Inocula
Preparation of the infectious inoculum involves growing the organism and measuring the number of infectious particles to be delivered. The organism is maintained as a frozen stock until required. Most single cell pathogens, including bacteria and yeasts, can be grown in liquid media. For some organisms, such as the yeast Cryptococcus, it is necessary to agitate the broth to maintain a single cell suspension and prevent clumping. For more complex pathogens, such as molds, it is necessary to wash or scrape the surface of an established culture on solid media to procure a suspension of infectious particles.
The gold standard for measurement of the concentration of organisms in suspension is serial dilution and overnight culture on solid media, followed by calculation of the concentration by counting the number of colony-forming units (CFUsl). The absorbency of the suspension at a given wavelength (usually 600 nm) can then be correlated with the concentration of organisms, as quantified by colony counts. By measuring the absorbency over a wide range of concentrations, one can construct a standard curve as a rapid and accurate estimate of concentration of organisms in a broth. For pathogens with larger infectious particles, such as fungal spores, the concentration can be measured directly using manual or electronic counting chambers. In addition, it is useful to measure the actual concentration of organisms delivered to the animals by serial dilution and colony counts.
To mimic human pneumonia, the delivery of the pathogen to the lungs is best achieved by the respiratory route. This route of administration requires that the mouse be anesthetized with an acceptable injectable anesthetic agent or anesthetic cocktail. In our laboratory, injection with either a dissociative anesthetic agent, such as ketamine, or a barbiturate is used. Once an adequate plane of anesthesia is reached, the inoculum can be delivered either by the intranasal or intratracheal route. The simplest approach is intranasal delivery. This technique is limited by the variability of the dose of pathogen delivered to the lungs, since the animal may swallow or exhale part of the inoculum, depending on the depth of anesthesia. Direct intratracheal delivery of the inoculum allows for more reproducible conditions. In this method, the trachea of the anesthetized animal is exposed via a small incision in the neck, and the inoculum is administered via a sterile 26-gauge hypodermic needle. In mice, a volume of up to 50 gl can be safely administered by this route without the risk of respiratory embarrassment. The intratracheal delivery of 102 to 103 CFUs of Klebsiella pneumoniae to CD-l, CBA/J, or BALB/c mice results in the development of impressive lobar consolidation, manifest by the intralveolar accumulation of neutrophils by 24 to 48 hr, with mononuclear phagocyte influx at later time points after bacterial challenge (Figure 1). Animals challenged with sufficient quantities of K. pneumoniae (>102 CFUs) will predictably develop bacteremia and death as early as 48 to 72 hr after challenge, depending on the initial bacterial inoculum.
The CD-1, BALB/c, and CBA/J mouse strains are particularly susceptible to the development of Klebsiella pneumonia, whereas C57BL/6 mice are more resistant to intra-pulmonary challenge with K. pneumoniae. The mechanism underlying these strain differences is incompletely understood in bacterial pneumonia. One possible mechanism is the strain-specific response to lipopolysaccharide, a component of the Gram-negative bacterial cell wall. The genetic basis of this difference has recently been recognized (Poltorak and others 1998). A variety of mechanisms have been found to account for strain-related differences in susceptibility to other infections. For example, in a Cryptococcus neoformans pneumonia model, C57BL/6 mice are more susceptible to progressive cryptococcal disease than CBA/J or BALB/c mice. In this infection, susceptibility to infection has been associated with a Th-2, rather than a Th- 1, cytokine response. Thus, strain differences must be taken into account when establishing an animal model of pulmonary infection.
Models of Specific Clinical Settings
The murine models of pneumonia can be altered in many ways to mimic particular clinical circumstances. For example, alcoholic patients are known to be particularly susceptible to pneumonia due to K. pneumoniae (Adams and Jordan 1984). To study the aberrant pulmonary host defense in this setting, we developed a murine model of chronic alcohol consumption by including increasing concentrations of ethanol in a calorie-controlled liquid diet over 2 wk (Standiford and Danforth 1997; Zisman and others 1998), resulting in blood ethanol levels of 120 mg/dL. In this model, alcohol-fed mice are markedly more susceptible to common pulmonary bacterial pathogens, particularly K. pneumoniae.
In general, immunocompetent humans and animals do not develop pneumonia when exposed to Gram-negative organisms, such as Pseudomonas aeruginosa. However, specific patient populations, such as critically ill patients with sepsis syndrome, are predisposed to the development of Pseudomonas pneumonia. To reproduce this common clinical scenario using an animal model, we induced sepsis in mice by performing cecal ligation and 26-gauge needle puncture (a well-characterized model of sepsis), then administered Pseudomonas aeruginosa intratracheally 24 hr later (Steinhauser and others 1999). In animals undergoing sham abdominal surgery, P. aeruginosa was rapidly and completely cleared from the lung within 24 hr. In contrast, mice undergoing cecal ligation and puncture readily developed pneumonia, P. aeruginosa bacteremia, and death as early as 48 hr from the time of Pseudomonas challenge. We and others have attempted to replicate airways infection that occurs in patients with bronchiectasis, a chronic infection of the airways often due to P. aeruginosa, by incorporating P. aeruginosa in agar beads that lodge in the airways and form a nidus of infection (O'Reilly 1995; Pedersen and others 1990).
Measurements of Outcome in Pneumonia
Methods of assessing the host response in pneumonia can be considered under two headings: those that quantify the extent of microbial invasion and those that provide insight into the cellular responses of the host. It is necessary to study both elements in combination to best understand host-pathogen interactions in pneumonia.
Measurements of Severity of Infection
Although survival studies are less than ideal, they are, in our opinion, a necessary and important outcome measurement when studying life-threatening diseases such as pneumonia. A limitation of survival as an endpoint is that the mechanism of death is unknown without further study. The microscopic appearance of the lung can also be used as an endpoint of severity of pneumonia but is limited by providing a qualitative, rather than a quantitative, measurement.
A second set of outcome measurements, the number of organisms in the lung, represents the balance between the multiplication of the organism on the one hand and its clearance by the host on the other. In infections with single-cell pathogens, such as bacteria and yeasts, the number of organisms in the lung at a given time can be measured by homogenizing the organ in a specified volume of saline, followed by serial dilution and culture of the slurry. This approach can also be used to quantify the dissemination of the pathogen. For example, determinations of blood culture CFUs in bacterial pneumonias and brain CFUs in C. neoformans pneumonia are often performed as measures of dissemination of the infections. Quantification of cultures of homogenized organs is not a reliable and valid method for some infectious organisms, such as molds and filamentous bacteria, which do not form distinct infectious particles in tissue. In pneumonia due to molds, such as Aspergillus species, an assay for chitin can be used as an indirect measure of burden of organisms (Lehmann and White 1975). Chitin is a complex carbohydrate found exclusively in the wall of fungal hyphae (the invasive fungal form) but not in fungal spores or mammalian tissue.
Measurements of the Host's Response to Infection
The absolute number of cells within the lung can be accurately measured by examining lungs digested by a collagenase enzyme and converted to a single-cell suspension. This method can, for example, be used to measure the dynamic influx of various cell types into the lung after challenge with an organism and to study the behavior of these cells in vitro; however, it does not provide any information about the anatomical location of the cells within the lung. Morphometric methods use geometric models to derive quantitative information from lung histology and can be used to measure both the number and location of cells within the lung (Weibel and Cruz-Orive 1997).
Bronchoalveolar lavage (BAL1) is another technique that allows sampling of the alveolar space of euthanized animals to estimate the magnitude of cell influx into airspaces. BAL is achieved by the instillation of phosphate-buffered saline into the lungs of mice via a polypropylene tube secured in the trachea, followed by collection of the fluid and cells for further study. It is important to note that this method results in overrepresentation of alveolar cells and fluid from normal alveoli, since these alveolar units are more distensible than units with significant airspace consolidation. Moreover, BAL does not allow for sampling of the pulmonary intersti-tium. Consequently, the number of cells obtained by this method substantially underestimates the total number of cells in the lung (Downey and others 1993). Cells procured by either BAL or lung digestion can be counted and characterized morphologically by microscopy of cytospin preparations, or identified by their cell surface marker expression using flow cytometry.
An indirect method of quantifying the presence of neutro-phils in the whole lung is by using an assay for myelo-peroxidase in organ homogenates. The level of this enzyme has been shown to correlate with intrapulmonary granulocytes (Goldblum and others 1985). This assay may include neutrophils that are firmly adherent to the pulmonary vascular endothelium and have not been detached by perfusion of the pulmonary vasculature. The assay may also overestimate the number of intact pulmonary neutrophils by measuring the enzyme released by lysed neutrophils. Given the unique limitations of each of the methods to assess the inflammatory response, we generally incorporate several of the techniques described for a particular study.
Finally, to assess the generation of mediators produced in the context of pneumonia, lung levels of cytokines and chemokines can be measured by reverse transcriptase polymerase chain reaction or Northern blot analysis for mRNA detection and enzyme-linked immunosorbent assay or Western blot analysis for protein detection. Enzyme-linked immunosorbent assay techniques can be applied to organ homogenates or fluid obtained by BAL, whereas reverse transcriptase polymerase chain reaction, Northern blot, or Western blot can be used with suspensions of cells isolated from BAL or lung homogenates. In addition, cells isolated from lavage or lung digestion can be cultured in vitro to study their behavior.
Methods Used to Establish the Role of Specific Cytokines and Cells in Pneumonia
Although the measurements of bacterial invasion and host response to the infection provide some insight into host-pathogen interactions, they provide limited insight into the causal role of specific cells and cytokines in the patho-physiology of pneumonia. To establish causal relationships, it is necessary to determine the effects of specific immune interventions on outcome in experimental models utilized, which can be achieved by selective depletion of specific molecules or cells (with antibodies or inhibitors) or by selective augmentation of molecules in either a systemic or compartmentalized fashion. We discuss the methods used for these purposes below.
Antibody-mediated Depletion of Cells and Mediators
Polyclonal antimurine antibodies can be generated by immunization of foreign species (such as rabbits or goats) against a murine molecule, administered in combination with an immunoadjuvant (such as Freund's), and collecting serum from the immunized animal. Animals can also be immunized against a synthetic polypeptide containing particularly im-munogenic portions of the target molecule. Serum from each immunized animal must be tested for the presence of antibodies and whether the antibodies are neutralizing using an in vitro assay for the molecule in question. Polyclonal antibodies are inherently variable because they are produced by multiple clones of plasma cells and by multiple animals: Batches of immune serum from different animals, or even from the same animal over time, can differ in their specificity for various epitopes on the target molecule and their neutralizing ability. These limitations can be circumvented by developing hybridomas that produce a monoclonal antibody against the target molecule. However, the neutralizing capability of monoclonal antibody tends to be inferior to that of polyclonal antibodies due to the recognition of a more limited number of epitopes.
Purified neutralizing antibodies or antiserum can be injected intraperitoneally in mice for systemic depletion of a specific molecule, with purified antibody from nonimmune serum or serum from a nonimmunized animal serving as a control. For example, in a study of the role of tumor necrosis factor-a (TNF1) in invasive Aspergillus pneumonia, neutralization of TNF using polyclonal antibodies led to reduced survival and increased lung burden of Aspergillus as measured by chitin levels (Mehrad and others 1999). We have also utilized a TNF soluble receptor:Fc immunoglobulin construct to deplete TNF in vivo, which has the advantage of 50- to 1000-fold greater neutralizing capability compared with anti-TNF antibodies. Antibodies directed to specific cell surface markers can also be used to eliminate selected cell populations. For example, a monoclonal antibody known to recognize the Ly-6G antigen on the surface of mouse granulocytes (Fleming and others 1993; Pennline and others 1990) can produce in vivo depletion of these cells. Similar antibody studies can target CD4+ and CD8+ T-cells as well as natural killer cells. All extrinsic antibodies can, themselves, be recognized as foreign proteins by the recipient animal, resulting in a type III (arthus-type) immune response. In practice, such antibodies can be administered repeatedly at 48-hr intervals to ensure neutralization but should be used with caution for periods longer than 7 to 10 days. Direct intratracheal delivery of antibodies to the lung has been used successfully to deplete TNF but in our hands has proven to be less efficacious than systemic antibody administration.
Depletion of specific cell populations can also be achieved pharmacologically, albeit with the disadvantage of influencing more cell populations than those targeted for depletion. Cyclophosphamide has been utilized to render rodents and other species transiently neutropenic; however, its use is complicated by its broad and nonspecific effects on various macrophage and lymphocyte populations. Depletion of macrophages in vivo has been more problematic. To address this issue, we administered dichloromethylene diphosphonate-liposomal complexes intratracheally to mice. This administration resulted in selective depletion of alveolar macrophages as early as 24 hr after administration of the complexes, which lasted for at least 5 days (Broug-Holub and others 1997). Depletion of alveolar macrophages using this technique markedly enhances susceptibility to intra-tracheally administered K. pneumoniae, despite a vigorous recruitment of neutrophils.
Administration/Transient Expression of Mediator Molecules
The simplest means of modulating the host response is by systemic administration of a mediator molecule. This approach is limited by several considerations. Cytokine mediators often display pleotropic and diverse properties depending on their concentration and whether they are expressed locally or systemically. For instance, TNF acts as a critically important proximal signal for the innate immune response when expressed locally. However, the systemic administration of TNF results in fever, vasodilatation, depressed myocardial contractility, and generalized activation of the coagulation cascade, thus mimicking features observed in sepsis syndrome. In addition, most mediator molecules are rapidly degraded and have a very short half-life when administered systemically. Lastly, the mode of action of some mediators, such as those involved in recruitment of inflammatory cells, depends on the establishment of a concentration gradient, which is difficult to achieve by systemic administration.
Some of these difficulties can be circumvented by site-directed administration of cytokines directly into the organ of interest. For example, in studies to elucidate the role of TNF in pneumonia due to K. pneumoniae and Aspergillus fumigatus, we have administered TNF or a TNF agonist peptide intratracheally, thereby limiting the extrapulmonary effects of the molecule (Laichalk and others 1998; Mehrad and others 1999).
Another method of delivery of cytokines in a compartmentalized fashion is by transient transgenic cytokine expression using intratracheal adenoviral gene therapy. Using this technique, replication-deficient recombinant human type 5 adenoviral vectors incorporating the gene of interest are administered intratracheally, resulting in transfer and expression of selected genes by host cells. Adenoviruses have a natural tropism for respiratory epithelium, and recombinant adenoviral gene products are typically expressed in large amounts, albeit transiently, for a period of several days to weeks. These characteristics make recombinant adenoviruses particularly attractive for studying the effects of short-term expression of cytokines mediators on outcome in pneumonia. Greenberger and others (1996) have demonstrated that the intratracheal administration of a recombinant adenovirus incorporating the interleukin-12 (IL1-12) p35 and p40 cDNAs concomitant with K. pneumoniae administration resulted in a marked reduction in Klebsiella-induced mortality. We have similarly shown improvements in bacterial clearance and survival when transiently expressing TNF within the lung at the time of Klebsiella challenge (Standiford and others 1999).
A notable limitation of recombinant adenoviruses as vectors for overexpression of mediators is the antigenic properties of the virus itself. Deletion of viral coding sequences in the viral vector can reduce its antigenicity and extend the expression of the gene of interest (Morsy and others 1998). In addition, the adenoviral vector itself appears to have detrimental effects on innate immune responses to bacterial pathogens, which we have observed when adenovirus is administered to the lung at doses of 5 x 108 plaque forming units or greater.
Use of Transgenic and Gene Knockout Mice
Genetic manipulation of experimental animals was made possible by parallel advances in recombinant DNA technology as well as development of techniques that allow hormonal control of reproduction and in vitro manipulation of eggs. Transgenic animals express a constructed foreign gene, the transgene, in specific organs, and knockout animals are animals in which both copies of a given intrinsic gene are rendered nonfunctional. Such genetically manipulated hosts can provide valuable insight into the role of specific inflammatory mediators in the setting of infection.
Knockout animals that lack the gene for a specific mediator or its receptor can be used to examine the consequences of the absence of the molecule of interest. Unlike antibody depletion studies, knockout animals are unaffected by considerations of the ratio of the antibody to mediator and immune response directed at the injected antibody. For example, we examined the role of lipid mediator products of the enzyme 5-1ipoxygenase in pneumonia by assessing the outcome of K. pneumoniae pneumonia in 5-1ipoxygenase knockout mice (Bailie and others 1996). Compared with wild-type animals, knockout animals displayed impaired bacterial clearance and increased mortality, which was associated with a decrease in the ability of alveolar macrophages to ingest and kill Gram-negative bacteria. Similarly, IL-6 and interferon-gamma knockout mice have been used to establish important roles of these cytokines in lung innate immunity against Streptococcus pneumoniae.
By using a promoter gene that is specific for lung tissue, transgenic animals can be designed to overexpress a particular mediator in the lungs only (Lira and others 1997). Lung-specific expression of the transgene can be further refined to a particular compartment of the lung. For example, use of a promoter region specific for Clara (bronchiolaf exocrine) cells (CC10) allows for expression of the transgene in the small airways, whereas the use of a surfactant apoprotein C promoter localizes the expression to type II alveolar epithelial cells within the airspace (Smith and others 1998). Transgenic animals provide a sustained and compartmentalized overproduction of the molecule of interest while avoiding the detrimental inflammatory/immunosuppressive effects of the vector delivery system, as is the case when using adenoviral gene transfer. We have used transgenic animals to assess the role of the C-X-C chemokine KC in host defense against K. pneumoniae pneumonia (Tsai and others 1998). C-X-C chemokines are small peptides that exert a variety of effects, including potent neutrophil chemotactic and activating properties. Expression of the KC transgene was localized to the lung by incorporating the promoter region of the CC 10 gene, which encodes a protein produced exclusively by Clara cells. The transgenic animals displayed elevated lung levels of KC at baseline, which was substantially upregulated upon challenge with K. pneumoniae. Importantly, KC transgenic animals inoculated with K. pneumoniae intratracheally had increased neutrophil recruitment, improved bacterial clearance, and increased survival compared with wild-type controls.
Use of transgenic and knockout animals is limited by several considerations. Production of a genetically altered host is a time-consuming process that sometimes results in a nonviable progeny. Alternatively, the progeny may have an unexpected phenotype. For example, TNF knockout mice have been found to lack splenic B cell follicles (Pasparakis and others 1996), and granulocyte macrophage colony-stimulating factor knockout mice develop progressive accumulation of surfactant hyaline material in the alveolar space, closely resembling the human disorder pulmonary alveolar proteinosis (Dranoff and others 1994). Furthermore, the deletion of a specific gene since conception may result in the development of altemative pathways to compensate for the absence of a specific molecule. Lastly, the missing molecule can result in an altered milieu, resulting in an aberrant response not directly attributable to the molecule of interest. Thus, results obtained using knockout animals can be quite disparate from those obtained using antibody depletion studies (Doerschuk and others 1996). In practice, it is advantageous to use studies in knockout animals in parallel with antibody depletion studies. Similarly, the chronic transgenic expression of specific molecule may result in attenuated host responses to the molecule of interest. A recently developed approach to this problem is the use of externally regulatable transgenic systems. An example of such an approach is the tetracycline-inducible system, whereby the CCI0 promoter is linked to a tetracycline-responsive regulatory element. Using these linked constructs, the gene of interest can be expressed only when doxycycline is added to drinking water. This system has been used to transiently express IL-11 in the lungs of mice, although such an approach has not yet been used in animal models of pneumonia (Ray and others 1997).
The use of neutralizing antibodies, alone or in combination with the use of knockout or transgenic mice, has provided much insight into the role of selected cytokines as mediators of leukocyte recruitment and/or activation in pneumonia. An overview of specific cytokine effects on leukocyte effector cell activities is presented in Table 1.
Application to Human Disease
In this article, we have reviewed the use of animal models in the study of host defense in pneumonia. An obvious goal of research in this area is identification of potential means of improving the host response to pathogens by selective augmentation or depletion of specific cytokine mediators in humans, thus providing an adjunct to conventional antibiotic therapy. This pursuit is limited by several factors. First, the pleotropic effects of individual cytokines can lead to unexpected consequences of such interventions in vivo. In addition, redundancies in the cytokine cascade can prevent a biologically significant effect if only a single cytokine is targeted. Furthermore, host responses with various host organisms can differ substantially. Finally, manipulation of cytokines in experimental animals occurs before or at the time of challenge with a specific pathogen, which is in contrast with the clinical setting, where patients usually present with the inflammatory cascade already in full swing.
Although the application of cytokine therapy can be problematic, several features of cytokine immunotherapy offer particular promise. Compartmentalized manipulation of cytokines can limit the extrapulmonary toxicities of cytokine therapy. Specifically, virally mediated gene therapy has emerged as a means of sustained localized expression of a specific cytokine mediator for a limited period of time (10 to 14 days). The redundancy of cytokine pathways may be partially overcome by agonists or antagonists directed at common receptors. Early intervention using such delivery approaches can be directed at persons at high risk of developing pneumonia, such as those who are critically ill, are on mechanical ventilation, or have prolonged granulocytopenia, with the goal of limiting the development or severity of pneumonia. For example, a recent randomized placebo-controlled trial examining the role of granulocyte colony-stimulating factor as an adjunct to antibiotics in patients with severe community-acquired pneumonia admitted to hospital showed a reduction in the incidence of serious complications, including acute respiratory distress syndrome and empyema, in patients receiving granulocyte colony-stimulating factor, indicating potential benefit in the most severely ill patients (Nelson and others 1998). In addition, given the inherent difficulties of treating pneumonia once established, the prevention of pneumonia in high-risk groups is of paramount importance. In this context, we have shown that the administration of a TNF agonist peptide administered intratracheally at least 3 days before bacterial challenge resulted in marked improvement in lung bacterial clearance and survival, compared with animals receiving control peptide intratracheally.
In conclusion, a better understanding of the host response to respiratory pathogens may allow for identification of specific immunotherapies as strategies to shift the balance between host defense and microbial virulence. Observations made in animal models of pneumonia have provided important insights that have brought us closer to identifying and implementing new approaches to therapy in the treatment of this devastating illness.
1 Abbreviations used in this paper: BAL, bronchoalveolar lavage; CFU, colony-forming unit; IL, interleukin; TNF, tumor necrosis factor.
Acknowledgments
This work was supported in part by National Institutes of Health grants HL57243, HL58200, and P50HL60289.
References
Adams HG, Jordan C. 1984. Infections in the alcoholic. Mad Clin N Am 68:179-200.
Anonymous. 1988. Statistical Abstract of the United States. Washington DC: US Department of Commerce, Bureau of Census.
Anonymous. 1997. Premature deaths, monthly mortality, and monthly physician contacts United States. MMWR 46:556-561.
Bailie MB, Standiford TJ, Laichalk LL, Coffey MJ, Striatar R, Peters-Golden M. 1996. Leukotriene-deficient mice manifest enhanced lethality from Klebsiella pneumonia in association with decreased alveolar macrophage phagocytic and bactericidal activities. J Immuno1157:5221-5224.
Barile MF, Grabowski MW, Kapatais-Zoumbos K, Brown B, Hu PC, Chandler DK. 1993. Experimentally induced Mycoplasma pneumoniae pneumonia in chimpanzees. Microb Pathog 15:243-253.
Baskerville A, Dowsett AB, Cook RW, Dennis MJ, Cranage MP, Greenaway PJ. 1991. Pneumocystis cariniipneumonia in simian immunodeficiency virus infection: Immunohistological and scanning and transmission electron microscopical studies. J Pathol 164:175-184.
Broug-Holub E, Toews GB, van lwaarden JF, Streiter RM, Kunkel SL Paine R 3rd, Standiford TJ. 1997. Alveolar macrophages are required for protective pulmonary defenses in murine Klebsiella pneumonia: Elimination of alveolar macrophages increases neutrophil recruitment but decreases bacterial clearance and survival. Infect Immun 65:1139-1146.
Caldwell A, Glauser FL, Smith WR, Hoshiko M, Morton ME. 1975. The effects of dehydration on the radiologic and pathologic appearance of experimental canine segmental pneumonia. Am Rev Respir Dis 112:651 -656.
Doern GV. 1995. Trends in antimicrobial susceptibility of bacterial pathogens of the respiratory tract. Am J Med 99:3S-7S.
Doerschuk CM, Quinlan WM, Doyle NA, Bullard DC, Vestweber D, Jones ML, Takei F, Ward PA, Beaudet AL. 1996. The role of P-selectin and ICAM-I in acute lung injury as determined using blocking antibodies and mutant mice. J Immunol 157:4609-4614.
Downey GP, Worthen GS, Henson PM, Hyde DM. 1993. Neutrophil sequestration and migration in localized pulmonary inflammation. Capillary localization and migration across the interalveolar septum. Am Rev Respir Dis 147:168-176.
Dranoff G, Crawford AD, Sadelain M, Ream B, Rashid A, Bronson RT, Dickersin GR, Bachurski C J, Mark EL, Whitsett IA, Mulligan RC. 1994. Involvement of granulocyte-macrophage colony-stimulating factor in pulmonary homeostasis. Science 264:713-716.
Fleming TI, Fleming ML, Malek TR. 1993. Selective expression of Ly-6G on myeloid lineage cells in mouse bone marrow. RB6-8C5 mAb to granulocyte-differentiation antigen (Gr-1) detects members of the Ly-6 family. J Immunol 151:2399-2408.
Garibaldi RA. 1985. Epidemiology of community-acquired respiratory tract infections in adults. Incidence, etiology, and impact. Am J Med 78:32-37.
Goldblum SE, Wu KM, Jay M. 1985. Lung myeloperoxidase as a measure of pulmonary leukostasis in rabbits. J Appl Physiol 59:1978-1985.
Greenberger MI, Kunkel SL, Strieter RM, Lukacs NW, Bramson J, Gauldie J, Graham FL, Hitt M, Danforth JM, Standiford TI. 1996. IL-12 gene therapy protects mice in lethal Klebsiella pneumonia. J Immunol 157:3006-3012.
Hanly P, Light RB. 1989. Plasma volume expansion and PEEP in a canine model of acute Pseudomonas pneumonia. Lung 167:285-299.
Holmberg SD, Solomon SL, Blake PA. 1987. Health and economic impacts of antimicrobial resistance. Rev Infect Dis 9:1065-1078.
Huxley EJ, Viroslav J, Gray WR, Pierce AK. 1978. Pharyngeal aspiration in normal adults and patients with depressed consciousness. Am J Med 64:564-568.
Jackson AE, Southern PM, Pierce AK, Fallis BD, Sanford JP. 1967. Pulmonary clearance of Gram-negative bacilli. J Lab Clin Med 69:833-841.
Johanson WG, Pierce AK, Sanford JP. 1969. Changing pharyngeal bacterial flora of hospitalized patients. Emergence of Gram-negative bacilli. N Engl J Med 281:1137-1140.
Johanson WG Jr, Pierce AK, Sanford JP, Thomas GD. 1972. Nosocomial respiratory infections with Gram-negative bacilli. The significance of colonization of the respiratory tract. Ann Intern Med 77:701-706.
Laichalk LL, Bucknell KA, Huffnagle GB, Wilkowski JM, Moore TA, Romanelli RI, Standiford TI. 1998. Intrapulmonary delivery of tumor necrosis factor agonist peptide augments host defense in murine Gram-negative bacterial pneumonia. Infect Immun 66:2822-2826.
Lehmann PF, White LO. 1975. Chitin assay used to demonstrate renal localization and cortisone-enhanced growth of Aspergillus fumigatus mycelium in mice. Infect Immun 12:987-992.
Lira SA, Fuentes ME, Strieter RM, Durham SK. 1997. Transgenic methods to study chemokine function in lung and central nervous system. Methods Enzymol 287:304-318.
Mehrad B, Strieter RM, Standiford TI. 1999. Role of TNF in pulmonary host defense in murine invasive aspergillosis. J Immunol 162:1633-1640.
Morsy MA, Gu M, Motzel S, Zhao J, Lin J, Su Q, Allen H, Franlin L, Parks RI, Graham FL, Kochanek S, Bett A J, Caskey CT. 1998. An adenoviral vector deleted for all viral coding sequences results in enhanced safety and extended expression of a leptin transgene. Proc Natl Acad Sci U S A 95:7866-7871.
Moser KM, Maurer J, Jassy L, Kremsdoff R, Konopka R, Shure D, Harrell IH. 1982. Sensitivity, specificity, and risk of diagnostic procedures in a canine model of Streptococcus pneumoniae pneumonia. Am Rev Respir Dis 125:436-442.
Nelson S, Belknap SM, Carlson RW, Dale D, DeBoisblanc B, Farkas S, Fotheringham N, Ho H, Marfie T, Movahhed H, Root R, Wilson J. 1998. A randomized controlled trial of filgrastim as an adjunct to antibiotics for treatment of hospitalized patients with community-acquired pneumonia. CAP Study Group. J Infect Dis 178:1075-1080.
O'Reilly T. 1995. Relevance of animal models for chronic bacterial airway infections in humans. Am J Respir Crit Care Med 151:2101-2107.
Pasparakis M, Alexopoulou L, Episkopou V, Kollias G. 1996. Immune and inflammatory responses in TNF alpha-deficient mice: A critical requirement for TNF alpha in the formation of primary B cell follicles, follicular dendritic cell networks and germinal centers, and in the maturation of the humoral immune response. J Exp Med 184:1397-1411.
Patterson GA, Todd TR. 1982. A large animal model of Pseudomonas pneumonia. J Surg Res 33:214-219.
Pedersen SS, Shand GH, Hansen BL, Hansen GN. 1990. Induction of experimental chronic Pseudomonas aeruginosa lung infection with P. aeruginosa entrapped in alginate microspheres. APMIS 98:203-211.
Pennline KJ, Pellerito F, DaFonseca M, Monahan P, Siegel MI, Smith SR. 1990. Flow cytometric analysis of recombinant murine GM-CSF (rmuGM-CSF) induced,changes in the distribution of specific cell populations in vivo. Cytometry 11:283-291.
Pinner RW, Teutsch SM, Simonsen L, Klug LA, Graber JM, Clarke MI, Berkelman RL. 1996. Trends in infectious diseases mortality in the Unites States. JAMA 275:189-193.
Poltorak A, He X, Smimova I, Liu MY, Huffel CV, Du X, Birdwell D, Alejos E, Silva M, Galanos C, Freudenberg M, Ricciardi-Castagnoli P, Layton B, Beutler B. 1998. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: Mutations in Tlr4 gene. Science 282:285-288.
Ray P, Tang W, Wang P, Homer R, Kuhn C, Flavell R, Elias J. 1997. Regulated overexpression of interleukin 11 in the lung. J Clin Invest 100:2501-2511.
Reynolds HY. 1997. Integrated host defense against infections. In: Crystal RG, West JB, Weibel ER, Barnes PI, editors. The Lung: Scientific Foundations. 2nd ed, vol 2. Philadelphia PA: Lippincott-Raven. p 2353-2365.
Sanders A, Crystal RG. 1997. Consequences to the lung of specific deficiencies in host defense. In: Crystal RG, West JB, Weibel ER, Barnes PI, editors. The Lung: Scientific Foundations. 2nd ed, vol 2. Philadelphia PA: Lippincott-Raven. p 2367-2379.
Smith S, Skerrett SI, Chi EY, Jonas M, Mohler K, Wilson CB. 1998. The locus of tumor necrosis factor-alpha action in lung inflammation. Am J Resp Cell Mol Biol 19:881-891.
Standiford TI, Danforth JM. 1997. Ethanol feeding inhibits proinflammatory cytokine expression from murine alveolar macrophages ex vivo. Alcohol Clin Exp Res 21:1212-1217.
Standiford TI, Wilkowski JM, Sisson TH, Hattori M, Mehrad B, Bucknell KA, Moore TA. 1999. Intrapulmonary tumor necrosis factor gene therapy increases bacterial clearance and survival in murine Gram-negative pneumonia. Hum Gene Ther 10:899-909.
Steinhauser ML, Hogaboam CM, Kunkel SL, Lukacs NW, Strieter RM, Standiford TI. 1999. lnterleukin 10 is a major mediator of sepsis-induced impairment in lung antimicrobial host defense. J Immunol 162:392-399.
Toews GB. 1994. Pulmonary clearance of infectious agents. In: Pennington JE, editor. Respiratory Infections--Diagnosis and Managment. 3rd ed. New York NY: Raven Press. p 43-53.
Tsai WC, Strieter RM, Wilkowski JM, Bucknell KA, Burdick MD, Lira SA, Standiford TI. 1998. Lung-specific transgenic expression of KC enhances resistance to Klebsiella pneumoniae in mice. J lmmunol 161:2435-2440.
Weibel ER, Cruz-Orive LM. 1997. Morphometric methods. In: Crystal RG, West JB, Weibel ER, Barnes PI, editors. The Lung: Scientific Foundations. 2nd ed, vol 2. Philadelphia PA: Lippincott-Raven. p 333-344.
Zisman DA, Strieter RM, Kunkel SL, Tsai WC, Wilkowski JM, Bucknell KA, Standiford TI. 1998. Ethanol feeding impairs innate immunity and alters the expression of Thl- and Th2-phenotype cytokines in murine Klebsiella pneumonia. Alcohol Clin Exp Res 22:621-627.
Table 1 Effect and relative potency of cytokines on leukocyte effector cell activities in murine Gram-negative bacterial pneumonia
| Cytokine | Biologic Effect | ||
| Leukocyte recruitment | Leukocyte activation | Regulation of cytokine | |
| TNFa | +++ a,b | +++ | +++ |
| G-CSFa | ++ | ++ | |
| GM-CSFa | ++ | ++ | + |
| Chemokines | +++ | ++ | + c |
| IFN-ga | +++ | ++ | |
| ILa-12 | +++ | ||
| IL-10 | _a | _ |
b indirect effect via induction of adhesion molecules and chemotactic cytokines.
c Specific to C-C chemokines.
Figure 1 Composite histologic secuons and bronchoalveolar (BAL) differentials from saline or Klebsiella pneumoniae-challenged mice. Panels A and B are histologic sections of lung 48 hr after the intratracheal administration of saline or K. pneumoniae (103 colony-forming units), respectively. Panels C and D are BAL differentials from K. pneumoniae-challenged animals at 48 hr at 40x and 100x, respectively.
Copyright © 2008. National Academy of Sciences.
All rights reserved.
500 Fifth St. N.W., Washington, D.C. 20001.
Terms of Use and Privacy Statement